Free AI agent agency course

Build AI-agent services that hold up in the real world.

Learn to choose the right workflow, design a bounded agent, connect data and tools, evaluate behavior, launch a controlled pilot, sell the work honestly, and operate an agency.

Start with lesson 1See the full curriculum
lessons
40
modules
10
free
$0
no account needed
Open

Your progress

Pick up where you left off.

Progress is saved on this device. You do not need an account.

0 of 40 lessons complete

Start the course

What this course changes

Build the service around a real job and a named owner.

A polished demo can hide a weak workflow. This course starts with the operating problem, then works through context, actions, approval, evaluation, launch, commercial scope, and ongoing responsibility.

Choose the smallest useful system

Separate rules, assistants, agents, and human decisions so AI is used only where bounded judgment has a purpose.

Design for failure before launch

Map hard paths, permissions, approvals, data authority, evaluation cases, stop conditions, incidents, and human handoff.

Sell a result you can deliver

Use discovery, a proof-based demo, explicit scope, pricing, onboarding, and a controlled pilot instead of hype.

Operate across clients

Standardize the parts that repeat while keeping client policy, data, access, quality, usage, margin, and support visible.

Complete curriculum

Forty lessons from the first workflow to a defensible agency service.

Every lesson produces a decision, artifact, or review rule. The capstone joins them into an offer and pilot plan a skeptical operator can inspect.

Module 01

Start with the job

Define agents carefully, model the economics, find the bottleneck, and decide whether AI belongs.

4 lessons0/4 completeStart this module
  1. Lesson 01

    What an AI agent is — and is not

    Choose the smallest useful system for the work instead of calling every automation an agent.

    Read lessonComplete
  2. Lesson 02

    The economics of a useful agent

    Measure an agent by the business constraint it improves, not by how impressive the demo looks.

    Read lessonComplete
  3. Lesson 03

    Find the workflow bottleneck

    Map the real journey from trigger to outcome before trying to automate the visible symptom.

    Read lessonComplete
  4. Lesson 04

    Decide whether AI belongs in the workflow

    Use an explicit scorecard before investing in an agent build.

    Read lessonComplete

Module checkpoint

Is the start with the job work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Define agents carefully, model the economics, find the bottleneck, and decide whether AI belongs.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 02

Find a market and package the work

Choose a reachable segment, research the workflow, shape the offer, and price ongoing responsibility.

4 lessons0/4 completeStart this module
  1. Lesson 05

    Pick a market without guessing

    Choose a reachable segment with a repeated operational pain and an identifiable buyer.

    Read lessonComplete
  2. Lesson 06

    Research conversations, not just competitors

    Turn customer language and workflow evidence into a useful problem inventory.

    Read lessonComplete
  3. Lesson 07

    Turn a use case into an offer

    Package a believable before-and-after workflow with scope, responsibilities, and proof.

    Read lessonComplete
  4. Lesson 08

    Price for delivery and ongoing responsibility

    Build transparent pricing around the work, usage, support, and risk you actually carry.

    Read lessonComplete

Module checkpoint

Is the find a market and package the work work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Choose a reachable segment, research the workflow, shape the offer, and price ongoing responsibility.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 03

Design an agent people can trust

Give it one job, map normal and hard paths, write auditable instructions, and design approvals and handoffs.

4 lessons0/4 completeStart this module
  1. Lesson 09

    Give the agent one accountable job

    Write an agent contract that makes its purpose, limits, and owner inspectable.

    Read lessonComplete
  2. Lesson 10

    Map the happy path and the hard paths

    Design the normal journey and the exceptions before real customers discover them for you.

    Read lessonComplete
  3. Lesson 11

    Write instructions people can audit

    Turn policies and SOPs into clear, testable instructions rather than one oversized prompt.

    Read lessonComplete
  4. Lesson 12

    Design tools, approvals, and handoffs

    Give an agent only the actions and information the job truly needs.

    Read lessonComplete

Module checkpoint

Is the design an agent people can trust work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Give it one job, map normal and hard paths, write auditable instructions, and design approvals and handoffs.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 04

Build the first useful workflow

Set the workspace and source of truth, capture intent, choose the next action, and preserve context.

4 lessons0/4 completeStart this module
  1. Lesson 13

    Set up the client workspace and source of truth

    Make ownership, lifecycle, and authoritative records clear before connecting an agent.

    Read lessonComplete
  2. Lesson 14

    Capture intent from forms, funnels, and conversations

    Turn a new signal into enough useful context for a person or workflow to act.

    Read lessonComplete
  3. Lesson 15

    Build the next-best-action workflow

    Make one narrow workflow that qualifies, records context, alerts an owner, and begins approved follow-up.

    Read lessonComplete
  4. Lesson 16

    Make the handoff feel continuous

    Preserve the customer’s context when a person takes over.

    Read lessonComplete

Module checkpoint

Is the build the first useful workflow work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Set the workspace and source of truth, capture intent, choose the next action, and preserve context.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 05

Serve people over voice and messaging

Design reception, qualification, booking, follow-up, and difficult conversations with an easy human route.

4 lessons0/4 completeStart this module
  1. Lesson 17

    Design an AI receptionist that helps, not blocks

    Create a concise inbound experience that identifies purpose, gathers permitted details, and routes to a person when needed.

    Read lessonComplete
  2. Lesson 18

    Qualify, route, and book with context

    Ask only approved questions, use transparent criteria, and preserve a clear human route.

    Read lessonComplete
  3. Lesson 19

    Use messaging for timely, welcome follow-up

    Make follow-up useful, permission-aware, and easy to stop or hand to a person.

    Read lessonComplete
  4. Lesson 20

    Handle difficult conversations

    Design for uncertainty, complaints, sensitive requests, and emergencies before launch.

    Read lessonComplete

Module checkpoint

Is the serve people over voice and messaging work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Design reception, qualification, booking, follow-up, and difficult conversations with an easy human route.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 06

Connect data and systems safely

Map data authority, connect APIs, build cited knowledge, and apply least-privilege permissions.

4 lessons0/4 completeStart this module
  1. Lesson 21

    Build a source-of-truth map

    Document where each fact starts, which system owns it, and how a correction flows back.

    Read lessonComplete
  2. Lesson 22

    Connect APIs and webhooks with guardrails

    Treat integrations as operational systems with retries, audit trails, and safe failure states.

    Read lessonComplete
  3. Lesson 23

    Build a knowledge base the agent can cite

    Curate approved, current information with owners and retrieval boundaries.

    Read lessonComplete
  4. Lesson 24

    Protect data and permissions

    Apply least privilege, tenant separation, retention choices, and regular access review.

    Read lessonComplete

Module checkpoint

Is the connect data and systems safely work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Map data authority, connect APIs, build cited knowledge, and apply least-privilege permissions.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 07

Test, monitor, and improve

Build evaluation sets, measure outcomes, layer safeguards, and run incidents and change control.

4 lessons0/4 completeStart this module
  1. Lesson 25

    Build an evaluation set before launch

    Test representative normal, edge, adversarial, and failure cases against expected outcomes.

    Read lessonComplete
  2. Lesson 26

    Measure the outcome, not only the conversation

    Track completion, quality, handoff, correction, and customer experience together.

    Read lessonComplete
  3. Lesson 27

    Operate with layered safeguards

    Combine permission, input, tool, approval, logging, and human safeguards instead of relying on one prompt.

    Read lessonComplete
  4. Lesson 28

    Run incidents, change control, and continuous improvement

    Release changes deliberately and learn from failures without hiding them.

    Read lessonComplete

Module checkpoint

Is the test, monitor, and improve work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Build evaluation sets, measure outcomes, layer safeguards, and run incidents and change control.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 08

Sell the implementation honestly

Create a believable demo, find conversations without spam, run discovery, and sell clear scope.

4 lessons0/4 completeStart this module
  1. Lesson 29

    Build a proof-based demo

    Show one believable workflow with assumptions, fallback, and visible operational value.

    Read lessonComplete
  2. Lesson 30

    Find conversations without spam

    Create an ethical acquisition system using relevance, permission, proof, referrals, partnerships, and useful education.

    Read lessonComplete
  3. Lesson 31

    Run discovery that produces an architecture

    Use discovery to understand the current workflow, decision rights, systems, risks, and measurement before proposing a build.

    Read lessonComplete
  4. Lesson 32

    Sell scope, not hype

    Present the delivery phases, shared responsibilities, exclusions, risk controls, and success measures plainly.

    Read lessonComplete

Module checkpoint

Is the sell the implementation honestly work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Create a believable demo, find conversations without spam, run discovery, and sell clear scope.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 09

Deliver client and white-label work

Onboard cleanly, launch a bounded pilot, set client expectations, and run evidence-led reviews.

4 lessons0/4 completeStart this module
  1. Lesson 33

    Onboard a client without losing context

    Collect the people, evidence, access, policy, baseline, and approvals required for a controlled build.

    Read lessonComplete
  2. Lesson 34

    Deliver a controlled pilot

    Launch with a bounded audience, stop conditions, daily review, and a clear decision point.

    Read lessonComplete
  3. Lesson 35

    Build a white-label client experience

    Create branded workspaces with separated permissions, clear support, and honest service expectations.

    Read lessonComplete
  4. Lesson 36

    Run client reviews that keep the work useful

    Use evidence-led reviews to discuss outcomes, exceptions, changes, and next decisions.

    Read lessonComplete

Module checkpoint

Is the deliver client and white-label work work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Onboard cleanly, launch a bounded pilot, set client expectations, and run evidence-led reviews.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Module 10

Scale a responsible agency

Template recurring work, manage margin and capacity, set team access, and defend the complete service.

4 lessons0/4 completeStart this module
  1. Lesson 37

    Standardize the 20 percent that repeats

    Turn recurring delivery work into reusable templates while preserving client-specific judgement.

    Read lessonComplete
  2. Lesson 38

    Manage margins, capacity, and usage

    Understand labor, provider usage, support, and service limits before scaling delivery.

    Read lessonComplete
  3. Lesson 39

    Build a delivery team and partner network

    Define roles and access so strategy, implementation, QA, support, and specialists can work without broad production permissions.

    Read lessonComplete
  4. Lesson 40

    Capstone: launch and defend an agent service

    Bring the course together in a complete, evidence-led offer, architecture, pilot, and operating plan.

    Read lessonComplete

Module checkpoint

Is the scale a responsible agency work ready?

The team finished the lessons, but the handoff only lists tasks. Choose the strongest response before moving on.

  1. Add more tasks so the handoff looks complete.
  2. Present one evidence-backed artifact showing how the team addressed: Template recurring work, manage margin and capacity, set team access, and defend the complete service.
  3. Start the next module and document this one at the end.
Reveal the reasoning

Answer: B. A list proves activity; a reviewable artifact proves judgment. Before moving on, the evidence should show the decision, its source, its limits, and the next step.

Course project

Launch and defend one agent service.

The capstone is not a pitch deck. It is the operating case for a narrow service: who it helps, what it may do, what can go wrong, how you will know, and who owns every decision.

  1. 01A workflow and agent job contract
  2. 02A source-of-truth, tool, permission, and handoff map
  3. 03An evaluation set with success, quality, and safety checks
  4. 04A scoped offer, price, onboarding, and controlled pilot
  5. 05A 90-day operating plan for reviews, incidents, usage, and change

How to finish it

Turn the pieces into a handoff someone else can review.

  1. 01Assemble all five deliverables and link every important claim to its source.
  2. 02Run a first review with the rubric. Mark assumptions and decisions that still lack support.
  3. 03Ask another person to challenge the plan without your explanation. Note where they get lost or ask for proof.
  4. 04Revise the handoff, assign owners and dates, and record what would cause the plan to change.

Defense rubric

A submission is ready when it survives these four questions.

Evidence
Can a reviewer find the source, date, and limits behind every important claim?
Judgment
Does the work explain why each decision was made and which alternative was rejected?
Operation
Is it clear who does what, when they do it, and what happens when something fails?
Review
Is there a measure, threshold, and date that tell the team whether to keep, change, or stop the plan?
See the 10-minute defense outline

Do not read the document from beginning to end. Present the reasoning in this order:

  1. The real problem and the evidence showing it deserves attention.
  2. The main decision, the trade-off it creates, and the alternative you rejected.
  3. The riskiest part of the plan and the protection you put around it.
  4. The first signal you will review, when you will review it, and what decision it will trigger.

Questions, answered

What to know before you begin

Do I need to be a developer?

No. The course explains the system in operational terms and includes tools, APIs, webhooks, knowledge, evaluation, and security. Technical readers can go deeper, but the decisions are written for founders, operators, and agencies too.

What counts as an AI agent in this course?

A bounded system that can use context to choose among allowed actions for one accountable job, then report what it did. A fixed rule is automation; a draft that a person approves is an assistant.

Does the course cover voice agents?

Yes. One module covers reception, qualification, routing, booking, messaging, difficult conversations, consent, urgency, sensitive requests, and human escalation.

How does it handle safety and security?

The course covers permissions, source authority, tool allowlists, approvals, data handling, evaluation sets, adversarial and failure cases, layered safeguards, monitoring, incidents, rollback, and change control.

Will this teach me to sell AI-agent services?

Yes. It covers market selection, research, offer design, pricing, proof-based demos, ethical acquisition, discovery, proposals, onboarding, pilots, white-label delivery, reviews, margin, capacity, and teams.

What do I build for the capstone?

A defensible agent service with a narrow workflow, owner, context and tools, permissions, evaluation set, safeguards, commercial scope, pilot plan, and 90-day operating review.

Next step

Build the operating layer around the agent.

Use Spacebrain to connect intake, CRM context, calls, messages, routing, tasks, automations, approvals, and reporting.

Start for free